perfSONAR User Q&A and Other Discussion

[Daniel Spisak <>]

I seem to be rather confused as to how to deploy perfsonar here. I am trying to do the following:

- Deploy a private centralmanagement server that is also a MA

- Deploy a network of testpoints

- Use a central configuration file

- Ensure maddash is functional

Do I need to setup lsregistrationdaemon.conf on any of my hosts if I want it to all be private?

In the testpoints, based on the documentation as it stands I am being told to create a meshconfig-agent.conf file with just the following content:

<mesh>
    configuration_url             http://172.28.128.5/perfsonar-mesh.json
</mesh>

However, in the Mesh Agent Config File reference there is no mention of the <mesh> primitive so I am wondering.

Moving onto MadDash, I was able to get that to load on my test setup, but it had a config that I had not created in it and it wasnt working. Looking at the error message for maddash it offered a number of things to verify. One of these was 

"Verify that /usr/lib/perfsonar/bin/generate_gui_configuration has run recently and you are looking at an accurate test mesh"

Except this binary does not exist at all on my centralmanagement server. What is this called now? 

Other suggestions it has are to check "/etc/init.d/perfsonar-regulartesting status" which also does not exist on the centralmanagement server. I assume this uses pScheduler now?

If there exists a clearly documented example of the kind of configuration of perfsonar I'm seeking I'd love some pointers as I'm currently struggling to get this working at all.

--

Daniel Spisak | Senior Dev Ops Engineer, Synthetic
AppDynamics
Monitor your Apps in Production. Now.
Watch our Video | Download AppDynamics for Free
Twitter | Facebook | appdynamics.com